7shifts is an HR platform designed to help restaurant teams thrive. With an easy-to-use app and industry-specific solutions, 7shifts saves time, reduces errors, and helps keep costs in check for over 55,000 restaurants. Our mission is to simplify team management and improve performance for restaurants, with a long-term vision of creating a thriving restaurant industry through the power of connected & engaged teams.

As the AI Data Architect at 7shifts, you'll design and build how the company asks questions of its data. This is a hands-on architect role: you write code, you ship, and you run the project. Reporting to the Head of Data, you'll partner with data and analytics engineers who own our governed, certified datasets and semantic layer, the IT team that runs our systems, and the Finance, GTM, and Product teams who need answers. Your first big outcome is an agentic self-service portal for basic data requests, live and adopted across those teams, to enhance the experience of business users and to enable our analysts to move to harder problems.

At a company our size, the architect and the builder are the same person. You'll define the end-to-end architecture for agentic data solutions, build the parts that need building, help refine the semantic layer alongside analytics engineering, own the business context layer beside it, and hold the line on what counts as a certified answer. You've shipped an agentic data solution before, watched it fail in places, and fixed it and you can run a complex, cross-functional project to a shared outcome with teams that don't report to you.

AI is both what you build and how you work. You'll use Claude and MCP daily,



and you'll be a thought leader on how to effectively use AI in the context of data across the organization.

What you’ll do:

- Lead the hands-on design of end-to-end agentic data solution architectures: how a question travels from a Finance, GTM, or Product user through the portal, the agent, the business context layer, and the semantic layer to a certified answer, including monitoring and evaluation of answer quality

- Ship the agentic self-service portal for basic data requests and drive adoption, measured by ticket volume, stakeholder feedback, and answer accuracy

- Own the front end: evaluate and select the portals and tools people use to interact with agentic data solutions, and partner with IT to integrate and secure them (identity, access, change management)

- Support the development of the business context layer: the definitions, business rules, and operational knowledge that let an agent interpret a question the way the business means it, kept current as the business changes

- Partner with analytics engineering on the semantic layer and certified datasets: they own the build, you help build and refine, and you bring the business context on what people need to ask and why

- Develop and maintain the technical roadmap for AI-on-data capabilities, and make the call on when a solution is positive enough to switch on even when governance isn't perfect

- Evaluate tools, platforms,



and frameworks (Claude, MCP, agent frameworks, retrieval and text-to-SQL approaches, Snowflake and dbt capabilities) through proofs of concept, weighing cost alongside quality, and recommend what we adopt

- Run AI-and-data projects end to end across data engineering, analytics engineering, IT, and business teams: discovery, requirements, test plans, rollout

- Advise business leaders on where agentic data solutions fit, and lead AI use cases across the organization

What you bring:

- A background in data engineering, analytics engineering, software development, or a related systems role (BI and analytics paths count too), with hands-on Python and SQL

- You’ve shipped an AI agent or LLM-based solution that answers questions from structured data, watched it fail, and fixed it. You can explain what went wrong and what you changed

- You build AI solutions with cost in mind. You’ve worked within an LLM budget, know what drives usage and spend, and can explain the trade-offs you made between answer quality and cost

- A working knowledge of how modern data stacks (dbt, Snowflake, or similar) are built and run, and of core data ideas: modelling, governance, certified data, and lineage

- A track record of leading complex projects across several teams from start to finish, keeping everyone pointed at the same goal

- An understanding of how businesses use data to run day-to-day, and experience turning that into something a system can use, like glossaries, business rules, or metric definitions

- Insight into what makes people use or ignore data tools (chat interfaces, AI built into reporting tools, internal portals),

📌 AI Data Architect (Toronto)
🏢 7shifts
📍 Toronto